Adrian Reid explains how discretionary traders can convert past, diverse trading approaches into a systematic, testable trading system. He advises isolating individual decision-rules and indicators from past strategies, codifying each as separate strategies in backtesting software (e.g., AmiBroker), analyzing trade logs to characterize decision triggers and exits, testing combinations and filters, discarding losing approaches, and keeping/refining only consistently profitable rule-sets.
Chapters
So if you've been trading before with lots of different methods, how can you take that and design a system out of it so that you can become more consistent? And I've done this quite a few times. The trick is to look at how you've made your trading decisions and as much as possible, isolate the different types of rules that you've used. So you might have, you might have had one strategy where you just followed this indicator. You know, maybe it was a moving average and you bought on crossover above the moving average and you sold when it crossed below the moving average or something, I don't know. Just for an example, identify the individual strategies that you used and codify them into your backtesting software. Put them into amibroker and back test them. Don't try and take all of your strategies and mush them into one system and hope that it works. What you want to do is take all of this, all of the rules, all the indicators that you've used in the past and try them out in combination to see which combinations are profitable and what you probably, what you usually find is if you're a break even trader, most of the strategies or indicators you've used in the past are probably losing you money. And there might be one or two ideas that do make you money consistently. What you're looking to do is to separate all of the approaches into separate ideas. Test them all systematically, throw away the bulk that doesn't work and keep the rules that do work and then turn that into a viable system.
If you're not sure yet, spend some time analyzing your trade log. You should be recording every trade, particularly if you're a discretionary trader. Analyze how you made the decisions and try and characterize those decisions. About okay, I use these indicators and I use this exit and, and I held for this long. Characterize them and then turn that into a system set of rules and test it to see if it's got an edge. If it's profitable, then you can maybe refine that, optimize it and make it better. Test some extra filters. But if it's at the beginning, if it just doesn't make money, just throw that approach away, don't use it again and find the different approaches or strategies that you've used in the past that do make money and focus on those. But you've got to get clear on how you're making the decisions and take those rules, codify them and test them. That's how you take a discretionary portfolio and systemize it.
